Parser edits for hpo, ctd

[colleenXu]

Main changes are:

• fixing the chunking issue for CTD disease-chemical
• only doing direct relationships for CTD disease-chemical
• merging records for CTD disease-pathway when the only difference is the inference_gene_symbol. To match what is shown in CTD like http://ctdbase.org/detail.go?type=disease&acc=MESH%3aD002375&view=pathway

[colleenXu]

Note from biothings 2/16 call:

• I suggested updating CTD data with this bug-fixed disease-chemical function soon. Since the current API is missing a huge chunk of the data
• Need more discussion with Kevin about the HPO references column.
  • It sounds like we don't want to convert the curies (OMIM:####, ORPHA:####, DECIPHER:####) to website urls since it would make the backend data larger
  • Instead, converting curies to website URLs is more of a frontend/UI issue

Proposal I will make to Kevin:

• A way to deal with this while helping downstream BTE is to split references into 6 separate fields by value type: ISBN, PMID, websites, OMIM, ORPHANET, DECIPHER.

[colleenXu]

Also examples of reference display issues in HPO website:

• not showing the decipher references correctly
• not showing the websites or ISBN references correctly

So the references that do display correctly are....

• OMIM references
• PMID references
• ORPHA (ORPHANET) references

[colleenXu]

Latest commit matches the proposal I wrote in the comment above